Simulation of anisotropic chemical etching of crystalline silicon using a cellular automata model

doi:10.1016/0924-4247(94)00820-5

Sensors and Actuators A: Physical

Volume 45, Issue 1, October 1994, Pages 85-89

https://doi.org/10.1016/0924-4247(94)00820-5 Get rights and content

Abstract

Anisotropic chemical etching of crystalline silicon in aqueous KOH is simulated at the atomic level using a cellular automata model. Experimental etch-rate ratios as well as the influence of temperature and concentration of the etchant are taken into account by introducing a stochastic component. With the help of two examples, the underetching of a convex mask corner and mask-corner compensation for mesa-type corners, the capabilities of this model are demonstrated.

References (19)

R.A. Buser et al.
ASEP: a CAD program for silicon anisotropic etching
Sensors and Actuators A
(1991)
J.S. Danel et al.
Anisotropic crystal etching - a simulation program
Sensors and Actuators A
(1992)
C.H. Séquin
Computer simulation of anisotropic crystal etching
Sensors and Actuators A
(1992)
M. Alavi et al.
Fabrication of microchannels by laser machining and anisotropic etching of silicon
Sensors and Actuators A
(1992)
S. Wolfram
Universality and complexity in cellular automata
Physica
(1984)
R.W. Gosper
Exploiting regularities in large cellular spaces
Physica
(1984)
K.E. Petersen
Silicon as a mechanical material
Proc. IEEE
(1982)
J. Frühauf, Simulation of orientation dependent etching of masked silicon substrates, Tech. Digest, Micromechanics...
J.N. Shepherd, Prediction of anisotropic etching in (100) silicon, Tech. Digest, Micromechanics Europe '90, Berlin,...

There are more references available in the full text version of this article.

Cited by (88)

2D Analysis and simulation of quartz crystal etch penetration by revisiting a previous geometric method
2024, Sensors and Actuators A: Physical
This paper presents a quartz wet etching simulation procedure tailored for microelectromechanical systems (MEMS) fabrication. It utilizes geometric methods and is primarily founded on modifying some aspects of the traditional slowness vector approach. Through a theoretical analysis of the crystal etching trajectory, we provide a macroscopic elucidation for the formation of 2D etch profiles and optimize wafer etch penetration during double-sided wet etching convergence of upper and lower surfaces. Based on the proposed theoretical model, we develop a CAD-like quartz crystal etch simulator that allows for simulations of various quartz cut types and etch penetration. To validate the accuracy of the simulator, we quantitatively compare the simulation with experimental results using the angle between adjacent etched sidewall profiles as a metric. The comparison reveals a minimum error of only 2.2° and a maximum error of 11.7° between the simulation and experimental results. The developed simulator is able to predict the profiles of different etching profiles well, and it has certain engineering application capability.
Etch and growth rates of GaN for surface orientations in the <0001> crystallographic zone: Step flow and terrace erosion/filling via the Continuous Cellular Automaton
2023, Materials Science in Semiconductor Processing
Citation Excerpt :
In order to describe the orientation dependence of the etch rate previous work by Zubel focused on the classification of the surface bonds instead of the surface atoms [21,22]. However, the resulting etch rates for crystalline silicon are exactly equal to those derived using a basic (or discrete) Cellular Automaton (BCA or DCA), which assigns the surface atoms with removal rates of value 1 (for most atoms) or 0 (for terrace monohydrides) [23–25]. Unfortunately, the BCA (and, thus, Zubel’s model) has very limited applicability (e.g., 30–40 wt% KOH on Si{100} wafers only) in comparison to the CCA, whose removal rates can take any value (not limited to 1 or 0), leading to successful simulations and modeling for many etchants, substrates and substrate orientations [18–20].
Compared to other methods, we present the benefits of the Continuous Cellular Automaton (CCA) to describe in a simple and flexible manner anisotropic wet chemical etching of GaN as a combination of step flow and terrace erosion. In fact, the simplicity of the approach enables accounting for epitaxial growth of GaN based on the equivalent perspective of step flow and terrace filling (or build-up). A key ingredient is the direct removal/deposition of GaN groups of atoms, which directly enables describing etching/growth via the removal/deposition of a small number of different crystallographic cell types, each with a different removal/deposition rate. We focus on the derivation of mathematical expressions for the etch rates of the surface orientations contained in the <0001> crystallographic zone as a function of the removal rates of the various cell types, then describing their use also for growth. A least squares approach is used to determine optimal values for the removal/deposition rates of the various cell types. For the first time, Focused Ion Beam (FIB) etching is used to manufacture a well-known, vertically micromachined wagon-wheel structure for subsequent use in wet etching, thus enabling a comparison between the CCA-derived etch rates and the experimental values. For growth, the comparison is performed against data from the literature. The combination of average step flow and average terrace erosion/filling, as inherently contained in the CCA method, explains well the anisotropy of both etching and growth of GaN in the <0001> crystallographic zone.
Review of atomic layer deposition process, application and modeling tools
2022, Materials Today: Proceedings
Citation Excerpt :
Only the primary direction rate is examined in the Ordinary-CA approach, which assumes only two fixed states of atoms on the lattice. Than and Buttgenbach [68] proposed the Random-CA approach. A probability parameter P determines whether the pre-deposited atom completed the opportunity procedure.
Atomic layer deposition is a highly sought-after technique with high conformal and quality film deposition. This study reviewed thin-film technology comprehensively and narrowed it to atomic layer deposition. Atomic layer deposition and associated terms and technology were discussed in detail. The study will be helpful to everyone interested in high precision and quality film deposition. The paper also discussed atomic layer deposition models and simulation tools with relevant studies examined. Molecular dynamics, Monte Carlo, and other models examined atomic layer deposition. Precision control of the sticking coefficients using Langmuir was also discussed. The paper will contribute to the body of knowledge for thin film, atomic layer deposition, modeling, and simulations with diverse real-life applications.
Modeling of silicon etching
2020, Handbook of Silicon Based MEMS Materials and Technologies
This chapter discusses the manufacturing and processing of MEMS structures. This consists in the sequential realization of material deposition, patterning and etching processes to incrementally modify the substrate and/or a multilayer structure in order to generate three-dimensional (3D) shapes with specific electromechanical functionalities. The set of micromachining processes has grown into a mature toolbox offering numerous alternatives depending on the application and material requirements. An overview of the currently available simulators is provided, centering mostly on the modeling of anisotropic etching due to the special challenges posed by this technique, as well as, its overall importance and generalizability. Major features of simulation methods of wet etching including commercial and free tools are discussed. The simulation of the different micromachining processes is possible only if a comprehensive database of different material properties and process parameters is available. Each micromachining process results in the 3D propagation of the material surface and/or interface. There are two types of atomistic simulator, depending on the actual method used for advancing time. A cellular automation is a discrete model of the physical system consisting of an n-dimensional lattice whose nodes, cells, sites, or atoms can be present in one of many states, where “many” can also be infinite. Although the calibration of the geometrical simulators is simpler there are several reasons to favor the use of the atomistic simulations. Various commercial and noncommercial simulators for modeling anisotropic etching, have been listed out.
Wet etching of silicon
2020, Handbook of Silicon Based MEMS Materials and Technologies
The use of an alkaline or acidic solution (the etchant) to chemically remove material from the surface of a sample is wet etching and this chapter explains the different types of wet etching. Although the etch rate anisotropy leads to faceting, the description of anisotropic etching as the propagation of a collection of crystallographic planes is an oversimplification. The important microscopic processes are briefly explained with relevant examples, which include pit nucleation and step flow, micromasking, diffusion, etc. The removal of the surface atoms during wet etching of silicon is a complex process that involves both chemical and electrochemical reactions. This chapter also provides an overview of the typical surface morphologies that can be observed during/after wet etching on different surface orientations, and an unified explanation of the mechanisms that may be responsible for it. The items used to characterize a silicon wafer like orientation issues, wafer dimensions, surface quality, and bulk features affect the etching result and thus silicon wafer plays a key role in wet chemical etching. This chapter gives detailed data on examples of wet etching, popular wet etchants, temperature dependence of the etch rate, concentration dependence of the etch rate and other variables affecting etch-rate values, etc. It also tries to give some data on the conversion between different measures of concentration.
A synchronous cellular automaton model of mass transport in porous media
2016, Computers and Chemical Engineering
Citation Excerpt :
The main issue thereby is to establish reasonable transition rules for a given process (Richards et al., 1990). For some mass transfer problems such rules have been formulated to study drug release (Zygourakis and Markenscoff, 1996; Bertrand et al., 2007; Laaksonen et al., 2009), diffusion of ligands across the protein surface (Kier et al., 2003), transport through a chromatographic column (Kier et al., 2000), chemical etching and corrosion (Than and Büttgenbach, 1994; Córdoba-Torres et al., 2001). Ideally, the set of rules should not explicitly describe every possible response of the model, i.e. not be phemonological.
In this work we present a fully synchronous coarse grained cellular automaton model for large-scale simulations at molecular level. The model is based on Margolus partitioning scheme, which was generalized as to describe quantitatively diffusion, adsorption and directed flow in porous media. Our aim is to create conceptually simple and computationally efficient framework to model the mass transport in porous materials with large representative volume. This work focuses on the fundamental aspects of the generalized Margolus cellular automaton. We exemplify the model by solving several diffusion problems, studying the monolayer adsorption, chromatography on disordered porous structures and chemical transformation in a system with phase separation. The results indicate that the model reflects the essential features of these phenomena. Absence of round-off errors, fully synchronous way of implementation, autonomous physically meaningful time scale and ease-to-handle boundary conditions make this model a promising framework for study various transport phenomena in porous structures.

View all citing articles on Scopus

View full text

Simulation of anisotropic chemical etching of crystalline silicon using a cellular automata model

Abstract

Sensors and Actuators A

Sensors and Actuators A

Sensors and Actuators A

Sensors and Actuators A

Physica

Physica

Silicon as a mechanical material

Proc. IEEE